Whole genome tiled arrays from Affymetrix, Agilent, and Nimblgen allow biologists to conduct unbiased chromatin immunoprecipitation coupled with DNA microarray analysis (ChlP-chip) to study the genome level in vivo binding of transcription factors (TFs). However, ChlP-chip on tiled arrays also generates massive amount data and poses a challenge for analysis algorithm development. We propose to develop effective computational algorithms to analyze ChlP-chip experiments on genome tiled arrays and model global transcription regulation mechanisms, for the goal of allowing biologists to adopt ChlP-chip to unravel the transcription regulatory network in mammalian genomes. Our specific aims are as follows. (1) Develop an open-source model-based algorithm to identify genomic regions enriched by TF ChlP-chip on Affymetrix tiled arrays. The approach will work with single ChlP-chip replicate without using mismatch probes or control experiments. (2) Analyze performance variability introduced in ChlP-chip procedures, tiled array platforms, and analysis methods. (3) Implement a public web server with integrated tools for sequence analysis of the global ChlP-regions in mammalian genomes. (4) Identify TF's cooperative binding partners and regulated genes, and model its global mechanisms of transcription regulation. Through highly collaborative efforts between computational and experimental biologists, including ChlP-chip, gene expression profiling, nucleosome occupation, and motif analyses, we will study estrogen receptor regulation in breast cancer cells. The proposed project will significantly enhance our understanding of global transcription regulatory mechanisms in mammalian genomes. Aim 4 will help answer the question what genes in breast carcinoma are regulated by estrogen receptor and why.